Model save

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6381
+- Loss: 1.2383
 
 ## Model description
 
@@ -49,10 +49,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.7095        | 0.09  | 1    | 1.7014          |
-| 1.7095        | 1.07  | 2    | 1.6855          |
-| 1.7095        | 2.06  | 3    | 1.6637          |
-| 1.7095        | 3.04  | 4    | 1.6382          |
+| 1.3801        | 0.35  | 4    | 1.3500          |
+| 1.3615        | 1.32  | 8    | 1.3000          |
+| 1.3064        | 2.3   | 12   | 1.2652          |
+| 1.2626        | 3.37  | 17   | 1.2383          |
 
 
 ### Framework versions

adapter_config.json

@@ -16,10 +16,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj",
     "k_proj",
-    "o_proj"
+    "o_proj",
+    "q_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM"
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0cbcd1726281528770049545897e4929fdbcd2e3c520f2cd8517683dc9413b7b
+oid sha256:733d047eede5712a63861b5f2821d0f11d45ce7cc9eef9931d98de25e00c1e80
 size 218138576

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 3.04,
-    "eval_loss": 1.6381348371505737,
-    "eval_runtime": 6.3457,
+    "epoch": 3.37,
+    "eval_loss": 1.2383112907409668,
+    "eval_runtime": 22.4776,
     "eval_samples": 294,
-    "eval_samples_per_second": 46.331,
-    "eval_steps_per_second": 5.831,
-    "train_loss": 1.9840126037597656,
-    "train_runtime": 1653.0192,
+    "eval_samples_per_second": 13.08,
+    "eval_steps_per_second": 1.646,
+    "train_loss": 1.30483017949497,
+    "train_runtime": 5957.4264,
     "train_samples": 5895,
-    "train_samples_per_second": 14.265,
-    "train_steps_per_second": 0.027
+    "train_samples_per_second": 3.958,
+    "train_steps_per_second": 0.007
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.04,
-    "eval_loss": 1.6381348371505737,
-    "eval_runtime": 6.3457,
+    "epoch": 3.37,
+    "eval_loss": 1.2383112907409668,
+    "eval_runtime": 22.4776,
     "eval_samples": 294,
-    "eval_samples_per_second": 46.331,
-    "eval_steps_per_second": 5.831
+    "eval_samples_per_second": 13.08,
+    "eval_steps_per_second": 1.646
 }

runs/Dec15_05-26-11_s4311/events.out.tfevents.1702618051.s4311.1026368.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17f33fcbcaf4a4ec5a6538ae3e5c63fe8a143b8ed099bf9b8607c390736e5aa8
+size 6281

runs/Dec15_05-26-11_s4311/events.out.tfevents.1702624031.s4311.1026368.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fa0b59f1d02d339bea48d497f4ff5acc85dc6a989e62ed82b0b704ead91ab41
+size 354

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.04,
-    "train_loss": 1.9840126037597656,
-    "train_runtime": 1653.0192,
+    "epoch": 3.37,
+    "train_loss": 1.30483017949497,
+    "train_runtime": 5957.4264,
     "train_samples": 5895,
-    "train_samples_per_second": 14.265,
-    "train_steps_per_second": 0.027
+    "train_samples_per_second": 3.958,
+    "train_steps_per_second": 0.007
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0400271370420624,
+  "epoch": 3.3670284938941655,
   "eval_steps": 500,
-  "global_step": 4,
+  "global_step": 17,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,56 +11,74 @@
     {
       "epoch": 0.09,
       "learning_rate": 1.9974521146102535e-05,
-      "loss": 1.7095,
+      "loss": 1.3801,
       "step": 1
     },
     {
-      "epoch": 0.09,
-      "eval_loss": 1.7013664245605469,
-      "eval_runtime": 6.3791,
-      "eval_samples_per_second": 46.088,
-      "eval_steps_per_second": 5.8,
-      "step": 1
+      "epoch": 0.35,
+      "eval_loss": 1.3500080108642578,
+      "eval_runtime": 22.5154,
+      "eval_samples_per_second": 13.058,
+      "eval_steps_per_second": 1.643,
+      "step": 4
     },
     {
-      "epoch": 1.07,
-      "eval_loss": 1.6854876279830933,
-      "eval_runtime": 6.3439,
-      "eval_samples_per_second": 46.344,
-      "eval_steps_per_second": 5.832,
-      "step": 2
+      "epoch": 1.06,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 1.3615,
+      "step": 5
     },
     {
-      "epoch": 2.06,
-      "eval_loss": 1.663727045059204,
-      "eval_runtime": 6.3388,
-      "eval_samples_per_second": 46.381,
-      "eval_steps_per_second": 5.837,
-      "step": 3
+      "epoch": 1.32,
+      "eval_loss": 1.2999709844589233,
+      "eval_runtime": 22.4738,
+      "eval_samples_per_second": 13.082,
+      "eval_steps_per_second": 1.646,
+      "step": 8
     },
     {
-      "epoch": 3.04,
-      "eval_loss": 1.6381714344024658,
-      "eval_runtime": 6.3614,
-      "eval_samples_per_second": 46.216,
-      "eval_steps_per_second": 5.816,
-      "step": 4
+      "epoch": 2.13,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 1.3064,
+      "step": 10
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 1.265210747718811,
+      "eval_runtime": 22.4153,
+      "eval_samples_per_second": 13.116,
+      "eval_steps_per_second": 1.651,
+      "step": 12
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 1.2626,
+      "step": 15
+    },
+    {
+      "epoch": 3.37,
+      "eval_loss": 1.2383277416229248,
+      "eval_runtime": 22.5157,
+      "eval_samples_per_second": 13.058,
+      "eval_steps_per_second": 1.643,
+      "step": 17
     },
     {
-      "epoch": 3.04,
-      "step": 4,
-      "total_flos": 2.123664876067881e+17,
-      "train_loss": 1.9840126037597656,
-      "train_runtime": 1653.0192,
-      "train_samples_per_second": 14.265,
-      "train_steps_per_second": 0.027
+      "epoch": 3.37,
+      "step": 17,
+      "total_flos": 7.670550714342441e+17,
+      "train_loss": 1.30483017949497,
+      "train_runtime": 5957.4264,
+      "train_samples_per_second": 3.958,
+      "train_steps_per_second": 0.007
     }
   ],
   "logging_steps": 5,
   "max_steps": 44,
   "num_train_epochs": 4,
   "save_steps": 500,
-  "total_flos": 2.123664876067881e+17,
+  "total_flos": 7.670550714342441e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b24e1f15eab76911260c95314f76d27bf8972d0e9bf4fdbcf3205fc1db47f9a
+oid sha256:6f1941f27569a238b088377cad71a086145d5d62cbf0cd926c9aae68936ed966
 size 4664